Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 12 Aug 1998 22:12:54 -0400
From:      Mike Tancsa <mike@sentex.net>
To:        questions@FreeBSD.ORG
Subject:   performance of a FreeBSD box: Can it keep up ?
Message-ID:  <3.0.5.32.19980812221254.00b083f0@sentex.net>

next in thread | raw e-mail | index | archive | help
I have been using a FreeBSD box as our internal router for the past while
and was just wondering how best to measure its capacity in that role.
Right now its a Pentium 200 with 64M of RAM with 3 intel fxp cards in it:
one via 100Mb cross over cable to a Cisco 4700 feeding us at 5Mb, the other
2 for 2 different segments.  Its managing everything via static routes
right now and I have a separate box runing gated to do OSPF internally for
my dynamic routing.  

My questions are as follows based on the netstat -s output below of the
router which has been up 111days and sees a fairly constant average of 1Mb
out and bursting to about 3.5Mb in.

a) Because in 111 days there have been only 9 source quenches does this
mean the box is keeping up ?
b) My only emperical tests that I have done are to do several big ftp
transfers from machines behind the box artificially pushing up the
utilzation to 4.6Mb for a good 5min period so MRTG can register it.  Are
there better ways to test its capacity
c) Am I better off from a performance point of view getting a Cisco for a
5Mb connection to the net that I am provisioned at.
d) Is the FreeBSD box's load average a useful indicator for this application
e) If I run gated and do BGP will it be able to handle the 5Mb feed + a T1
with the added processing requirements of router recalculation etc?  A
PII-333 with 256M of RAM is still WAY less expensive than a 4700, so I
could beef up the box.
f) Although I know ftp.cdrom.com of course can push out way larger numbers,
does its traffic patterns i.e. sending out just ftp vs. a typical ISP mix
make it harder/easier on a routing device.

Thanks for any insights/comments.


ip:
	3668374680 total packets received
	0 bad header checksums
	0 with size smaller than minimum
	0 with data size < data length
	0 with header length < data size
	0 with data length < header length
	99 with bad options
	0 with incorrect version number
	593 fragments received
	592 fragments dropped (dup or out of space)
	1 fragment dropped after timeout
	0 packets reassembled ok
	23875737 packets for this host
	40789 packets for unknown/unsupported protocol
	3630406314 packets forwarded
	2990494 packets not forwardable
	60428875 redirects sent
	20928778 packets sent from this host
	7616 packets sent with fabricated ip header
	0 output packets dropped due to no bufs, etc.
	0 output packets discarded due to no route
	0 output datagrams fragmented
	0 fragments created
	0 datagrams that can't be fragmented
icmp:
	5750035 calls to icmp_error
	19583 errors not generated 'cuz old message was icmp
	Output histogram:
		echo reply: 107130
		destination unreachable: 2438494
		routing redirect: 14548994
		time exceeded: 3270480
		parameter problem: 13
	0 messages with bad code fields
	0 messages < minimum length
	0 bad checksums
	0 messages with bad length
	Input histogram:
		echo reply: 484
		destination unreachable: 5043
		source quench: 9
		routing redirect: 48655
		echo: 107130
		time exceeded: 6969
	107130 message responses generated
tcp:
	98184 packets sent
		84024 data packets (7800244 bytes)
		451 data packets (162959 bytes) retransmitted
		0 resends initiated by MTU discovery
		6937 ack-only packets (5750 delayed)
		146 URG only packets
		729 window probe packets
		4902 window update packets
		996 control packets
	135836 packets received
		82911 acks (for 7836370 bytes)
		1397 duplicate acks
		0 acks for unsent data
		88538 packets (16867291 bytes) received in-sequence
		100 completely duplicate packets (894 bytes)
		0 old duplicate packets
		159 packets with some dup. data (282 bytes duped)
		606 out-of-order packets (49128 bytes)
		0 packets (0 bytes) of data after window
		0 window probes
		152 window update packets
		3 packets received after close
		12 discarded for bad checksums
		0 discarded for bad header offset fields
		0 discarded because packet too short
	433 connection requests
	371 connection accepts
	9 bad connection attempts
	0 listen queue overflows
	754 connections established (including accepts)
	814 connections closed (including 122 drops)
		177 connections updated cached RTT on close
		177 connections updated cached RTT variance on close
		83 connections updated cached ssthresh on close
	5 embryonic connections dropped
	81730 segments updated rtt (of 82222 attempts)
	424 retransmit timeouts
		0 connections dropped by rexmit timeout
	0 persist timeouts
		0 connections dropped by persist timeout
	562 keepalive timeouts
		556 keepalive probes sent
		6 connections dropped by keepalive
	7462 correct ACK header predictions
	47704 correct data packet header predictions
udp:
	23612399 datagrams received
	0 with incomplete header
	2 with bad data length field
	8 with bad checksum
	460530 dropped due to no socket
	23002098 broadcast/multicast datagrams dropped due to no socket
	0 dropped due to full socket buffers
	0 not for hashed pcb
	149761 delivered
	461749 datagrams output


	---Mike
**********************************************************************
Mike Tancsa, Network Admin        *  mike@sentex.net
Sentex Communications Corp,       *  http://www.sentex.net/mike
Cambridge, Ontario                *  01.519.651.3400
Canada                            *

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3.0.5.32.19980812221254.00b083f0>